ML Kit is now in GA & Introducing Selfie Segmentation
Blog post from Google Cloud
ML Kit has reached general availability, marking a significant milestone since its initial launch in 2018, with enhancements to its machine learning capabilities on both Android and iOS platforms. Among the new features is the Selfie Segmentation API, which allows developers to separate backgrounds from users in images and videos, facilitating the addition of creative effects in apps. The Pose Detection API has also been updated to include more poses, a 50% reduction in model size, and a new Z coordinate for depth analysis, enhancing its utility for fitness and yoga applications. Additionally, ML Kit now supports pose classification, developed in collaboration with the MediaPipe team, to help developers classify specific poses and count repetitions in exercises. These updates reflect ML Kit's commitment to providing cutting-edge tools for app developers, emphasizing the importance of community feedback in shaping its development.